Strategy — GORGE

GORGE has one number and it does three jobs. Almost every real decision in the game is a fight between those three jobs, and most losses are somebody resolving that fight badly.

Length is your score, your weapon, and your fuel — at the same time. It is what you are ranked on. It is the wall you cut people off with. And it is the only thing boost can be paid for with. Spending it to chase makes you shorter, which makes you worse at killing and lower on the board. There is no separate energy bar because the tension is the game.

Do not boost in the first thirty seconds

A fresh digger is barely above the floor where boosting stops being possible. Held down, a new digger's entire boost budget is gone in about half a second — long enough to feel like an input, nowhere near long enough to catch anybody.

So the opening is not a fight. It is shopping. Get length on the board first; boost becomes a real tool at roughly the point where you can hold it for several seconds and still have a body worth having afterwards. Players who arrive from twitch shooters lose their first dozen runs to boosting on instinct at a moment when boost cannot possibly pay off.

You cannot out-turn anyone, so stop trying

Every digger in the arena turns at the same fixed rate. There is no agility stat, no character that corners harder, and no input trick that tightens your arc. Two diggers in a circling contest will circle forever.

Which means pursuit is a geometry problem, not a reflex problem. Cut the chord. Do not steer at where a rival is; steer at the point their fixed turn rate forces them to arrive at. Because that rate is a known constant and identical for both of you, the intercept is genuinely computable in your head after a few runs, and it is the single skill that separates players who get kills from players who get near-misses.

Fresh rock is an anti-boost shield

This is the most useful thing on this page. Digging into un-excavated rock at normal speed is free. Entering it while boosting is fatal.

So when something faster is on you: drop your boost and dive into virgin ground. Your pursuer now has exactly two options, and both of them are good for you.

They keep boosting and follow you into the rock, and they die without you doing anything.
They drop boost to follow safely, and you are now both at the same speed with a gap between you — which is to say the chase is over.

The corollary is worth internalising too: an excavated area is fast and dangerous, and an unexcavated one is slow and safe. When you are winning a fight, keep it in the open tunnels. When you are losing one, go where the ground is solid.

Boost only through ground you have already seen

Following from the same rule: full speed through unfamiliar terrain is a way to kill yourself with no rival involved. A wall death with nobody in pursuit is credited to nobody, which is the game's polite way of saying it was entirely your fault.

Your own tunnel network is the ground you can safely run at speed, because you know it is clear. That is a real home-field advantage and it is worth building deliberately: a loop of your own tunnels near where the food is dense gives you an area you can fight in at double speed while everyone else has to be careful.

Tunnels persist for the life of the arena and they do not belong to whoever dug them. Kill someone and you inherit their map — their careful safe network becomes yours.

Pick your targets by reading tails

A dead digger leaves remains worth eating, and how much is there depends on how long they were. Tail composition exists so you can make that judgement across the arena instead of from a scoreboard: a rival's body visibly shows what they ate and where they have been.

A long, mineral-rich tail is a meal. It is also, usually, someone who has enough length to boost for a long time and enough body to cut you off with, so the meal is priced correctly. The read that actually pays is finding somebody long and low on options — deep in open excavated ground, far from fresh rock, with nowhere to shield.

Kills are not deliveries. The remains sit in the world and anyone can eat them. If you spent half your body boosting into a kill and a third digger is closer to the corpse than you are, you have just fed a rival at your own expense. Set up kills where you are the nearest mouth.

The head-on trade favours whoever has less

Two heads meeting kills both diggers, always. Length does not help. Nothing helps.

Read what that means competitively: a head-on is an even exchange in a game where the two players usually have very unequal stakes. If you are the longest digger in the arena, a head-on costs you a dominant position and costs a small rival almost nothing. If you are small, a head-on against the leader is one of the best trades available to you.

So: when you are ahead, never contest a straight line. Turn early, force the engagement to be side-on, and make them hit your body rather than your face. When you are behind and being hunted by someone twice your size, driving straight at them is a real threat and they know it.

Your risk appetite should invert at the lock

Arenas take joiners for a while and then lock. That single event changes what a death costs, and it should change how you play.

Before lock Dying is nearly free — you lose your length but you are in a new arena instantly, and the tunnels you carved were already banked to your country's explored total the moment you dug them. Take the aggressive line. A run that ends in a failed ambush cost you almost nothing.

After lock Nobody new arrives and nobody comes back. Every death is permanent removal, and the arena can only get smaller. Survival is now the whole game — this is the phase where the win exists, and it is the phase to stop taking even trades.

One detail people plan around wrongly: at the lock, any bots still in the field forfeit and leave — but they leave without dropping remains. There is no feast at the lock. Do not position for one.

Playing the collapse

After enough of the map has been dug out, the ground starts coming back: sections of tunnel cave in at random, and an outer edge closes on a clock. Both kill on contact. From the first cave-in you have roughly ninety seconds to two minutes.

Arrive with a bank. The finale is the one time you will genuinely need several seconds of boost in reserve, and it is the worst possible time to be at the floor. Length spent chasing in the mid-game is length you do not have here.
Walls come back, so cutting off works again. By the endgame the field has excavated most of the map, and wall-forcing has quietly stopped being available. The collapse hands walls back. If you have spent the round unable to trap anyone, the last two minutes are when that changes.
The closing edge is not a wall. It does not fill with rock; the ground out there simply stops being survivable. You cannot hide behind it and you cannot force somebody to crash into it — but you can force somebody to stay out there a moment too long.
Do not boost out of a tight pocket. A pocket with one exit and a rival in the corridor is survivable, because you can always dig a second exit — at normal speed, through rock, for free. Boosting at that wall is the failure mode the situation is designed to tempt you into.

Hiding works, right up until it doesn't

There is no starvation timer in GORGE, and that is a deliberate choice — a hunger mechanic would have punished aggression and rewarded sitting still, which is backwards for this game. So yes, a quiet corner is genuinely safe for a while.

The closing edge is what collects on that debt. It shrinks on a fixed schedule regardless of how anybody is playing, and it does not care that you are hidden. A player who camped arrives at the final space with no length, no boost, no tunnel network and no read on the survivors — against people who spent the round eating. The passive line does not lose to a rule. It loses to everybody else having got better while you waited.

Five ways good players still die

Boosting on reflex at the start, spending a budget that was never big enough to matter.
Chasing into unfamiliar ground at speed. The kill was two seconds away and the wall was one.
Winning a chase and losing the corpse to whoever was standing closer.
Accepting a head-on while leading. An even trade is a loss when you have more on the table.
Reaching the collapse empty. Every second of mid-game boost is a second you will not have in the finale.

New here? Start with How to Play →

The parts that are not obvious