Robot Battle Games: A Complete Guide

Robot battle games look like they are about reflexes and are actually about timing, which is a different thing. This guide covers the combat system underneath them - how attacks are structured, what a telegraph is telling you, why holding a guard is a losing strategy, and the specific habits that separate players who clear eight waves from players who clear three.

The three phases, in detail

Every attack in a well-built robot battle game - yours and the enemy's - is split into three phases. Understanding them is not optional background knowledge; it is the entire game.

The figures above are the actual values used by the games on this site: a light attack winds up in 0.05 seconds and recovers in 0.16, a heavy winds up in 0.24 and recovers in 0.34. They are quoted because the ratios matter more than the absolute numbers.

Notice that the recovery is always longer than the active window, usually by a factor of two or three. That asymmetry is deliberate and it is what makes committing to an attack a real decision rather than a free action. Every swing you take is a short moment of threat followed by a longer moment of vulnerability.

Notice also that a heavy attack's recovery is more than twice a light attack's. Heavy attacks are not simply better - they buy damage with time spent unable to defend, which is why they belong in openings rather than in openers.

Why attacking first loses

Here is the arithmetic that most new players never quite work out, and it explains the single most common way robot battles are lost.

Suppose an enemy begins a wind-up. You see movement and press attack. Your light attack winds up for 0.05 seconds and is active for 0.11, so you connect at around 0.16 seconds - and then you are in recovery until roughly 0.32 seconds. Their attack, winding up for 0.20, becomes active at 0.20 and stays active until 0.31. Your recovery sits exactly inside their active window.

So you trade. Both of you take a hit. That would be acceptable if the exchange were even, but it usually is not: most enemies in these games deal more damage per hit than you do, and you have a combo multiplier to lose that they do not.

The alternative is to wait. Let the wind-up finish, move out of the way or deflect, and attack into the recovery that follows. You take no damage and you land a hit against a target that cannot respond. In most robot battle games this single change accounts for more improvement than anything else available.

Reading telegraphs

A telegraph is the visual signal that an attack is coming. Good games draw it identically on every enemy so that what you learn from one transfers to all of them; this site uses a ring that closes around the attacker.

The colour carries the important information. Amber means the attack can be blocked or deflected. Red means it cannot, and the only answer is not to be there. Mixing those up is expensive: raising a plate against a red telegraph does nothing at all, and the time spent raising it is time not spent moving.

The other thing worth learning to read is telegraph length, which varies by enemy type. Heavy machines telegraph for noticeably longer than light ones. That is not just flavour - it means a heavy enemy is the safest thing on screen to punish and a light one is the most likely to catch you.

Deflecting versus blocking

These are usually the same button, and they are completely different actions with completely different value.

A deflect is a plate raised inside a short window before the hit lands - 0.18 seconds in the games here. It costs no power, staggers the attacker for nearly a second, and pays the largest defensive score bonus available. A block is a plate that was already up when the hit arrived. It stops the damage, but it costs a point of power every single time and contributes nothing offensively.

This is why holding the guard down permanently does not work, and why it is designed not to work. Hold it and every incoming attack becomes a block, every block drains power, and when power runs out the plate overloads and leaves you open for most of a second - typically right in the middle of a crowd.

The correct mental model is that the deflector is a tool you use at a moment, not a state you live in. Press it when the ring is about to close, not before.

One more thing that catches people: a deflector only covers a cone in front of you. Being flanked while blocking is a real failure state, not a bug, and it is why positioning matters even in a game where you are standing still.

Power management

Attacks, deflects and boosts usually draw from one shared meter that refills after a pause rather than continuously. That design has a specific consequence: a flurry of actions leaves you with nothing for several seconds, and several seconds is a very long time in a fight measured in tenths.

The discipline worth building is simple and unglamorous. Never spend your last point. A boost you cannot afford is the most common cause of taking a hit you could see coming, and in a game with a combo multiplier, taking a hit is usually the most expensive thing that can happen to you.

A useful way to think about it: treat the meter as having a floor at one point rather than at zero. Everything above that floor is spendable; the floor itself is your escape.

Positioning, and why standing still loses

In any battle game with more than one enemy, position is worth more than aim. A machine standing in the middle of an arena will eventually be surrounded, and once it is surrounded a front-facing deflector protects almost nothing.

The technique that solves this is called kiting: moving so that a crowd of pursuers strings out into a line rather than closing into a ring. It works because enemies move at different speeds, so a wide arc naturally sorts them by pace. You then fight the front of the line - one or two machines - instead of the whole crowd.

In practice that means moving in arcs rather than straight lines, and it means never backing into a corner. A corner removes two escape directions and gives everything on screen a reason to converge on the same point. If you find yourself in one, leave immediately rather than trying to fight clear.

This also explains a counter-intuitive targeting rule: the nearest enemy is usually not the most important one. The nearest is generally the fastest, which means it will still be near you in two seconds regardless. The one that matters is whatever will be sitting in your escape path when you get there.

What 'open' means, and why it is worth chasing

Most robot battle games have a state - variously called open, staggered or vulnerable - in which an enemy takes substantially more damage and pays a much larger score bonus. On this site an open target takes double damage and is worth roughly 1.6 times the points.

Enemies become open by being deflected, by being mid-wind-up when your attack lands, or by recovering from something. Learning to spot the state is probably the highest-value perceptual skill in the genre, because it changes which target you should be hitting moment to moment.

The practical rule: a staggered enemy across the arena is a better target than a healthy one next to you, provided you can reach it before the state expires. Almost nobody does this instinctively and almost everybody improves when they start.

A practice routine that works

If you want to get better at robot battle games specifically, this sequence is more efficient than simply playing more.

• Play one full run without attacking at all. Survive, watch, and count recovery lengths. You will lose, and you will learn what every enemy's telegraph looks like.
• Play a run where you only attack into recoveries. Your score will be poor and your survival will be much better than usual.
• Play a run where you deliberately deflect rather than dodge everything amber. This calibrates the deflect window, which is the hardest timing in the genre to feel.
• Play a run where you only move in arcs and never in straight lines, and notice how differently the crowd behaves.
• Then play normally. The four habits will be partly automatic and you will see where the remaining gap is.

Each of those takes a few minutes, because runs are short. The whole routine costs less than half an hour and is worth considerably more than half an hour of ordinary play.

Reading the seven enemy types

The battle games share one roster. Each type asks a different question, and recognising which question you are being asked is most of the skill.

Two of these are worth singling out. The turret is the only enemy that is more dangerous at range than up close, which inverts the usual instinct to keep distance. And the saw is the only one with no correct way to attack it - it is a moving piece of terrain, and treating it as an enemy is a mistake the game does not correct for you.

Why shielded machines are an angle problem

Shielded machines are where most intermediate players lose the most time, usually by treating them as a damage problem.

The plate sits on the machine's leading edge and only covers the direction it faces. Attacking into it does no damage at all - the hit registers as blocked, you get a metallic clang and a spray of sparks, and your attack's recovery happens anyway. Do that four times and you have spent two seconds achieving nothing while the rest of the wave arrived.

There are two correct answers. The first is positional: get to its flank. A shielded machine turns to face you, but it turns slowly, and moving in an arc around it will consistently produce an opening on its side.

The second is pressure, and it only applies where a game gives the machine a power meter of its own. Each blocked hit costs it power, and when that runs out the plate breaks and leaves it open for roughly a second. That is slower than flanking but does not require space, which makes it the answer when you are cornered.

The economy of a wave

It helps to think about a wave as a budget rather than a list of things to kill.

You have a fixed amount of power, a fixed number of hits you can absorb, and a combo multiplier that is worth more the longer it survives. Every action spends one of those. A shot spends power. A block spends power. A hit taken spends integrity and the multiplier together, which is why it is usually the most expensive thing that can happen.

The implication is that the cheapest way to clear a wave is rarely the fastest. Deflecting an attack costs nothing and produces an opening, so a deflect-then-punish exchange is strictly better than a dodge-then-punish, which is better again than trading. Players who clear waves quickly and then fail three waves later are usually paying for speed with resources they needed.

Frequently asked questions