In one fight, tracking them is difficult. In the other, they barely move across your screen.
The only thing that changed was your movement.
Which target would you rather track?
Throughout this guide, blue is you and red is the enemy. Orange means your crosshair is on target.
Your mouse is not the only thing controlling how hard a target is to track.
Your movement changes their movement on your screen.
This guide is about learning to control that relationship.
You are tracking someone smoothly. They change direction.
For a fraction of a second, your mouse keeps moving the old way. You have not reacted yet, so your aim is still following the movement you just saw. By the time you respond, the target has moved away from your crosshair.
Now you have to correct. The farther they moved during that delay, the larger the correction. We will call that distance the gap. It appears in yellow throughout this guide.
The figure separates the miss into two moments. First comes the delay before you respond to the reversal. Then comes the correction that brings your crosshair back to the target. Watch how your movement changes the second part even when the reaction delay stays the same.
You cannot remove the delay between seeing a reversal and responding to it. But your movement can change how quickly the gap grows while you are waiting to react. The same reaction delay can leave you with a large correction in one case and almost none in another.
Try it. Switch between standing still and the other options, and watch the yellow number. Your reaction time never changes. The correction you owe does.
Reaction delay. There is always some delay between the target reversing and your response beginning. The exact amount varies from player to player and from moment to moment.
Correction. Once you respond, you still have to close whatever gap opened during that delay. A larger gap usually means a larger correction. Your movement during the reaction delay already affects how large that correction will be.
You cannot eliminate reaction delay, but you can use movement to control how much correction is waiting for you when you respond.
So how does your movement change the gap? Start with the simplest possible case.
The enemy dodges left and right. You have two options: when they move, move in the same world direction, or in the opposite one.
Try both. Ignore accuracy for now. Watch only how fast they travel across your screen.
Move with them and their motion across your screen shrinks. You are both drifting the same way, so from where you stand they barely move. Move against them and their motion across your screen grows: your speed and theirs add together.
Your feet just changed how much work your mouse had to do. Nothing about the enemy changed at all.
These two choices have names. Moving in the same world direction as the enemy is called mirroring. Moving in the opposite world direction is called anti-mirroring.
The keys can feel backwards because you are facing each other. If both players press their own left-strafe key, they move in opposite world directions. Think about where each player is actually moving, not which key they pressed.
Because you are facing each other, the same strafe key sends the two players in opposite world directions. To move with an enemy who is facing you, you generally press the opposite-side strafe key.
In the last example, the timing was handled for you. Your direction changed exactly when the enemy changed theirs. Doing that yourself is the actual skill.
Strafe aiming takes the choice from the previous section and adds a timing rule: you change direction in response to the enemy’s direction changes rather than reversing on your own schedule.
To strafe aim is to change direction whenever, and only whenever, the target you are aiming at changes direction.
AIMER7, Strafe Aiming 101, §2.2
Notice that the rule describes your movement, not your mouse. It does not tell you how to move the crosshair. It only tells you when your movement direction changes.
There are two ways to break the rule. You can miss one of the enemy’s reversals, or you can reverse without the enemy reversing first. The second is not necessarily bad movement, and we will see several reasons to do it later, but it is no longer strafe aiming. Keeping those ideas separate makes both easier to practice.
Try it. Press ← and → below. Answer every change of theirs, and invent none of your own. It is harder than it sounds, and the score counts the two failures separately.
Strafe aiming uses your direction changes to make your own tracking more manageable. Those same direction changes could instead be used to make yourself harder to hit, or to reposition. That tradeoff matters more than whether a form is theoretically clean.
There is one important limit. If the enemy reverses faster than you can respond, you cannot reliably answer every change.
When that happens, stop chasing each reversal individually. Hold closer to the center of the pattern and let the target move back through your crosshair. A repeated short dodge becomes easier to anticipate because the player keeps returning through the same space.
That is also why one fast reversal inside a longer dodge can work so well. Repeating it gives away the pattern; using it once can catch the opponent before they adjust.
Drag the slider. Somewhere around a quarter of a second the better response changes. The exact crossover depends on the player, the target, and the situation. The slider is there to make the tradeoff visible, not to give you a universal timing threshold.
Why does moving with someone make them easier to track? You already know the answer from ordinary life.
Imagine someone running past you at 10 mph. Now start running beside them at 10 mph. Relative to you, they barely move. You could hold a conversation. Turn around and run the other way, and they fly past you.
Nothing about their speed changed. Your movement changed their speed relative to you. The same relative-motion idea applies to aiming, because your screen shows the world relative to you.
Your mouse does not care about all of that relative motion equally. The important part is the motion that crosses your line of sight, because that is what moves the target sideways across your screen.
Movement toward or away from you changes the range instead. That still matters: the same sideways speed covers much more of your screen up close than it does at long range.
Try every direction. Watch how your movement either cancels part of the enemy’s screen motion, adds to it, or lands somewhere between.
So far you have moved left and right. But you have more keys than that. What happens if you answer their dodge by moving forward, backward, or on a diagonal?
A pair of answers is called a form: one direction for when they go left, one for when they go right. Mirroring and anti-mirroring are the two you know. The other useful forms use diagonals. We will introduce them by what they feel like, when they can help, and what they trade away.
If both players move at 10 units per second, these come out as 0, 2.93, 10, 17.07 and 20 across your screen. The exact values are included for reference. You do not need to memorize them to use the idea.
What it feels like. The enemy nearly freezes on your screen. Neither of you has to move the mouse except at their direction changes.
Use it when you are happy to trade hits: more health, more damage per second, a teammate on the way.
Cost. It freezes you on their screen too. Because mirroring reduces the amount of lateral tracking for both players, it can reduce the value of having better raw tracking than the opponent.
What it feels like. In the simple left-right case, your movement and theirs add together, so the target crosses your screen much faster. Your mouse has to keep moving instead of settling.
Use it when you are comfortable turning the duel into a faster tracking test.
Cost. Both players have to aim hard. Being smooth and precise at this speed is genuinely difficult, especially up close.
What it feels like. Mirroring on a diagonal. The target can look almost stationary, but that does not necessarily make the shot effortless. Very slow motion exposes small precision errors, while the changing range keeps altering the mouse speed you need.
Use it when you want to reduce lateral screen speed while still changing range, especially when fine control is a strength.
Cost. Heavy on precision for both players. On a diagonal you also cover ground more slowly than their straight dodge, so you are a little easier to read than they are.
What it feels like. Fast lateral motion with a changing range, so the mouse speed you need is less constant than in a plain anti-mirror.
Use it when you want a fast tracking duel while also closing or opening distance.
Cost. The changing range makes the target’s screen speed less consistent, which can make the pattern harder to read for both players.
What it feels like. You move toward and away from the enemy rather than strafing across them. From a purely lateral-tracking perspective, it can look similar to standing still.
Use it when you have another reason to change range while keeping lateral mouse movement small, such as being safe behind a shield or out of range.
Cost. You give the opponent very little lateral movement to track, so this is usually a situational tool rather than a pattern to prioritize on its own.
Count the possibilities: eight directions you might be moving in when they turn, and seven you could switch to. That makes 56 possible response pairs. The five families above are the ones with exact opposites. The rest combine them, and a handful are genuinely useful.
They are cataloged in Going deeper. You do not need them yet.
Hold each form. Watch the mouse trace under the monitor, which is the work your hand is doing. Then swap the camera and see what the same form does to their screen. Every form changes the fight for both players, so always check the other camera too.
Back to the gap. The enemy changes direction, and you have not reacted yet, so for a brief moment your feet keep doing whatever they were doing.
During that instant, one of two things happens. Your movement can help close the mistake, or it can make the mistake larger.
If your current direction carries you the same way they are now going, your own movement partially cancels the target’s new motion on your screen, so they stay closer to where your crosshair already points. If it carries you the other way, every one of those milliseconds adds to the correction.
The figure runs that experiment for all eight keys at once. They reverse, you hold one key for one reaction time, nobody touches the mouse, and the bar is how far off them you end up. The dashed line is what standing still would have cost. Anything under it helped, anything over it hurt.
The original guide names these two cases. A direction change that reduces the coming correction is called inward; one that increases it is called outward. Every form you met in Part five is one or the other: mirroring and half-sideways mirroring are inward, both anti-mirrors are outward, and back-and-forth is one of each.
Slow your reaction time down and watch the bars stretch. The slower you notice, the more your choice of key matters.
Inward forms tend to leave you with a smaller correction after a reversal. Outward forms can leave you with a larger correction, but they also create more screen speed for both players. Which one you want depends on the kind of duel you are trying to create.
At this point it is tempting to ask which form is best. That question misses the tradeoff.
Mirroring can make both players nearly stationary on each other’s screens. Anti-mirroring does the opposite: both players have to track much faster. The diagonal forms create their own mixes of speed, precision, and changing range. None of these choices hands you extra accuracy on its own.
What a form decides is which skill the fight tests.
So the real question is: what kind of aim duel do I want this to become?
Change the mouse-control setting and compare how sensitive each form is to it. Do not read the bars as win probability. They show what this simplified model emphasizes when one variable changes. Mirroring changes least, because there is much less lateral motion to track.
Each row is the same twenty-second example under one form. Bars to the right show a modeled time-on-target advantage for you; bars to the left favor the opponent. This is not a prediction of who wins a real match. The table below gives the raw percentages.
If time favors you, with more health or help on the way, an even trade is a win. The mistake is trading when the only thing you had over them was aim.
So far we have looked at movement mainly through the lens of your own aim. In a real fight, your movement has other jobs too.
Sometimes your movement needs to help your aim. Sometimes it needs to make you harder to hit. Sometimes it needs to carry you behind a wall, toward your team, or out of another enemy’s line of fire. Good movement has a purpose. The important skill is knowing which job matters most right now.
Strafe aiming is one tool, not a rule you owe the whole fight. Use it when helping your aim is the priority. Break it when dodging or repositioning matters more, then return to it when the situation changes. Practicing the forms makes that switch easier, because you no longer have to think through every keypress.
One simple improvement is to make sure your dodge is taking you somewhere useful. Over a few seconds, drift one way more than the other. Do not finish the fight standing where you started it.
If your dodge stays centered on one spot, an opponent can hold near that center and let you cross their crosshair repeatedly. Add some drift and the center of your movement keeps changing, so that shortcut becomes less reliable. More importantly, the same drift can carry you toward cover, out of an angle, or back toward your team.
Set the drift to zero and watch what happens when the crosshair simply waits near the center of the dodge. Then add drift and see how quickly that fixed point stops being useful.
Movement usually has to trade among these three jobs. In a real fight, decide which one or two matter most right now.
Walls, corners and distance change how much room the enemy has and how much of them you can see. That is a whole subject of its own; see the companion guide to geometric positioning.
You now know enough to use strafe aiming in a real match. Respond to the enemy’s reversals. Decide whether you want to move with or against them. Keep your overall movement taking you somewhere useful. Everything below this point adds depth; none of it is required to start practicing the idea.
Everything from here is for readers who want the whole framework. There is a lot of it, and none of it is required to get value from the first eight parts.
Mirror left and right for several seconds and you finish where you started. The same is true of any basic form built from two opposite directions: one half cancels the other. Your tracking may be fine, but the pattern itself does not move you toward cover, your team, or a better angle.
The figure shows four seconds of three forms as tracks on the ground. The first two return to their starting point every cycle, forever. The third, built from a quarter turn instead of a half turn, goes somewhere.
So what if you want to keep responding to the enemy while also traveling somewhere?
Answer their left with forward-right and their right with forward-left. Now both halves of the pattern move you forward overall. The guide calls this a forward triangle. Reverse it and you retreat; rotate it and you begin to circle. None can be held forever, which is why they get combined.
Start with a left-right mirror. Now replace one of the two responses with forward-right. If you make that switch at a moment when both forms call for the same key, you can move from one pattern into the other without adding an extra reversal.
The guide calls that shared response a connection. Connections let you chain forms together so that your local dodge still responds to the enemy while the larger path carries you somewhere.
The basic opposite-direction forms do not connect directly to one another. Once quarter-turn forms are added, however, you can build longer chains that travel forward, backward, or around the opponent while your local dodge keeps responding to the enemy.
Build a chain. The figure only offers forms that legally connect to where you are, with the shared key written on the arrow. Try closing a rhombus, then try circling them.
Build a pattern that travels only toward or away from the enemy, through connections.
Build one rhombus in each direction of travel, and work out which side of each is the inward side.
Learning how much mouse movement each change of direction requires takes practice. Comparing two similar-looking patterns and noticing why they require different corrections is useful training even before you try to use the full chain in a match.
The forms so far react to what the enemy does. More advanced patterns try to anticipate how the enemy will respond to your movement. They are less about memorizing a shape and more about recognizing a tendency: show the opponent one problem, see how they solve it, then use that response to predict what comes next.
That is why they need timing and a feel for how players react when their aim is under pressure. The three below are the ones the original guide works through, each shown with the situation it is for. Notice the signature of all three: the enemy’s change of direction is predictable, and you are still, technically, strafe aiming.
Use movement to test what the opponent is comfortable with. If a pattern repeatedly makes them alter their own movement to stabilize their aim, that response gives you information. Once their response becomes consistent, their next direction change becomes easier to anticipate.
Eight directions you might be holding when they turn, seven you could switch to. The whole space fits in one table, and the named forms occupy only a small outlined subset of it. The number in each cell is how fast the enemy crosses your screen while you hold that pair.
Point at any cell. The diagonal is empty because staying on the same direction is not a change of direction, so it does not define a form.
Calm play becomes easier when the fight gives you fewer things to process at once. If your movement keeps the target manageable on your screen, your position removes some escape routes, and you engage on your own timing, the fight becomes easier to read.
Panic can create the opposite problem. If you spam movement keys, jump without a reason, and flick at every change, you make your own view faster and less predictable. Now you have to process some of the chaos you created yourself. Simple movement can make the fight easier to see.
The catch is that calm play does not feel like doing anything. Holding a long strafe, or keeping the mouse still at the center of their dodge, can be the best move available and still feel passive. Confidence in these plays only comes from having seen them work, which is what practice is for.
When you watch better players, do not copy isolated keypresses. Look for repeatable decisions: when they lengthen a strafe, when they reverse, when they let a target cross the crosshair, and when their movement starts drifting toward cover. Try to understand the purpose before copying the pattern.
Coordinating movement and aim takes deliberate practice. Two habits make that practice much more useful.
Practice one thing at a time. Choose one focus for the session: a specific form, adding drift, or moving toward cover while fighting. In one-versus-one modes, use the pause between rounds to choose a single focus for the next fight.
Practice somewhere experimentation is cheap. Customs, aim trainers, unranked modes, or any environment where you are willing to try movement that does not work yet. If you care too much about the immediate result, you will usually fall back to what already feels comfortable.
One habit worth stealing: bind an aim-bot to your fire key in an aim trainer’s free play, and move alongside it. The bot has perfect mouse control, so it shows you what the mouse movement for the form you are learning should feel like. This can be a useful way to build a feel for a new one.
A score can be useful feedback, but it may reward accuracy, survival, speed, or damage rather than the movement skill you are trying to train. Know what the score is measuring and do not let it become the goal of the drill. The comparison below shows what each system pays for.
These are three KovaaK’s routines adapted from the original guide, ordered from beginner to advanced, with a working timer. Start any block and the rest run in order. The note under each block explains the skill it is meant to train; that purpose matters more than the scenario name or score.
The lab lets you change the assumptions and see how the simplified model responds. Use it to compare relationships, not to predict the outcome of a real duel.
Watch two things: how large the modeled correction becomes after a reversal, and whether the setup favors one player’s time on target.
A smaller correction is not automatically an advantage. If the same form also makes the enemy’s aim easier, you may simply have made the fight easier for both players.
This glossary includes every term used on the page, including terms that the main guide deliberately introduces late. The original AIMER7 terminology comes first, with a plain-language description beside each entry.
The framework, the terminology, the form catalog and the routines come from Strafe Aiming 101, published free by AIMER7 in August 2019 as a companion to his earlier guide on geometric positioning. The material on purpose, drifting, control and calm draws on Fundamentals of Strafe by Sam (@400apm), July 2022, itself written to lead readers to the AIMER7 guides. Both are worth reading in full. This site adds interactive diagrams and a simplified movement-and-aim model to make some of those relationships easier to inspect. The model is not empirical evidence that one form is universally better in live play. It deliberately removes many things that matter in a real game so that one relationship can be examined at a time.
The interactive examples simplify the game on purpose. The eight movement directions use the same nominal movement speed and acceleration; the weapon behaves like a simple tracking weapon; and the examples ignore jumping, recoil, spread, reloads, projectiles, abilities, map geometry, and many other game-specific effects unless a section explicitly adds them.
The 10-units-per-second movement speed is only a convenient reference value inherited from the original guide. The useful relationships do not depend on memorizing that number.
These assumptions are useful for isolating the idea, not for reproducing a particular game.
Put the line of sight along x. You look along it, the enemy looks back along it, and their left hand points where your right hand points. A held key becomes a world vector by rotating it into that player’s frame. Split both velocities into a part across the sight line and a part along it; only the first is what the mouse must follow.
with the ground speed and each key’s angle from its own player’s forward. Because you face each other, the two contributions enter with opposite signs, which is why pressing the same key doubles the speed and pressing the opposite one cancels it. For the five families this gives 0, 2.93, 10, 17.07 and 20 exactly: mirroring cancels, anti-mirroring adds, a diagonal contributes , and back-and-forth contributes nothing across at all.
Angular rate is relative speed divided by range, so halving the range doubles everything. That is why a form that is comfortable across a room is unplayable at close quarters, and why the slow half-sideways mirror is the long-range pick.
Eight choices for the first half of a form and seven for the second give ordered pairs. Classified by the angle between your two directions: each direction has one opposite and two neighbours at each of 45°, 90° and 135°, giving 8 + 16 + 16 + 16. One caution: the original guide names two advanced forms “45-forms” after the corner of the triangle they draw on the ground; classified by the turn itself, the guide’s own metric everywhere else, they are 135° forms, and this site classifies by the turn throughout.
The classification is usually given as a rule with cases. This site measures it instead: freeze the crosshair at the instant of reversal, hold a candidate key for one reaction time, and record the angle between crosshair and body; standing still is the baseline. That one experiment reproduces every label in the guide, including back-and-forth coming out mixed. There the deciding factor is the range: pushing in shortens it, and screen speed goes as one over the range.
What this model can tell you: how the assumptions on this page interact mechanically. What it cannot tell you: which movement is optimal against a particular player, hero, weapon, map, team composition, or live-game situation.
Both simulated players run the same controller. It has two delays, and they are different things: tracking runs on visual data about 70 ms old, while noticing that the target’s rate has jumped is cognitive and slow. That second delay is the reaction time the sliders set. Between surprises the hand matches the rate it believes the target has, with a light trim. After a surprise the belief holds its old value for one reaction time while the gap opens at the full relative speed; then one ballistic flick closes it, aimed once, missing by a fixed fraction of its own size, so a bigger correction is also a worse one. Tracking wobble is a settling random walk whose size grows with commanded mouse speed, which is the only place “mouse control” enters the model.
Nothing in the controller knows what a form is. The forms are inputs, and the accuracy ranking the model produces (mirror, then half-sideways mirror, then back-and-forth, then the anti-mirrors) matches the ranking in the original guide. This is an internal sanity check on the implementation, not independent validation of the guide. What the model leaves out: walls and height, jumping, projectile travel, spread, and an enemy who reads you back (the simulated enemy is a dodge script with a period and a drift).
In a few cases, the simplified measurement used here gives a different answer from the original guide. Those cases are listed below so the assumptions and the disagreement are visible rather than hidden.
The automated tests make sure code changes do not alter the calculations unintentionally. They test the internal consistency of this site; they are not evidence that the simplified model captures every real-game situation.
npm test