hnA tacFPS Aim Guide (previously known as AimGud)
By: trgjtk #3598 with a special thanks to jdK, Prodigies, jake, Shaqary, and koopa for testing, feedback, and moral support.
Introduction:
The purpose of this document is to enlighten those who would like their aim to prosper. This is intended mostly for valorant or other tactical first person shooters/low time-to-kill games, but, in principle, at least using some of these thoughts can be productive for learning to aim in other games as well. Ultimately I intend to posit some theory to you, and provide relatively rigorous evidence that it is true, as well as give practical application of this theory.
First things first, MISCONCEPTIONS:
“Kovaak or Aim Lab, or whatever aim trainer you use is bad for improving at valorant.”
I really don’t care if you play gridshot, like someone who rides the streamer wave, or 1w6ts, like some mere sheep, the reality is, neither are useful for tacfps. As are most KovaaK scenarios, unless played correctly. There is essentially no correlation between scores and rank.
“The reason why aim trainers aren’t useful is NOT because these games don’t require aim.”
This is a terrible take that I see all the time and it honestly baffles me. Tenz for example; when he plays against other tier 1 pros, he completely destroys them. You cannot sit there with a straight face and say that there is not a clear aim difference.
At a certain point, the impact of other mechanics are more or less equalized between any player and their given opponent. Once you get to this point, the difference between aiming competency of two opposing players becomes increasingly obvious. This is also known as aim diff. Of course, this is assuming that you don’t actually have a complete lack of understanding of how other mechanics or the game as a whole works (admittedly a relatively weak assumption for many players *cough cough* plat/dia players with an ego).
Now you realize aim is relevant, and you realize that there is essentially minimal correlation between any kovaak scores and rank in game.
What does this mean?
So let’s see: You need to aim in valorant to be good unless you have a brain, which is admittedly a harder task. And yet “aim” trainers don't really do a good job of sharpening this skill. Why is that? Some people would say that this is due to aim trainers improving “mouse control” and not aim. I cannot fathom people who think this. Think about it, what is aim really?
It’s fundamentally made up of two components:
When do you aim without trying to control your mouse? You move the mouse from point A to B, as quickly and precisely as possible. And that’s aim. So with all that pretentious crap out of the way, let me define what I mean by Aim. Aim is how proficiently one is able to move their crosshair. This can be further broken down into game specific aim, as in, how proficient one is at moving their crosshair in the way the game they’re playing demands. For my purposes, this definition doesn’t include other mechanics, like crosshair placement, or game specific target acquisition (recognizing and aiming at the enemy).
So let’s summarize what I just covered:
This last point is obviously an important one to analyze because there appears to be a paradox in the idea that “aim trainers” aren’t actually improving aim in certain relevant aspects. So let’s analyze this further:
Why do we see that, empirically, people see better results with the use of aim trainers in some games more than others? It makes sense that the more specific the aiming task, or mouse movement, the better the training will transfer over in game.
A common trend that we observe is that aim trainers tend to be better for higher ttk games than lower ttk games. So somehow, the aim training done in Kovaak’s tends to be more similar to the aiming done in a higher ttk game compared to a game like Valorant. By nature, these higher ttk games also tend to be more tracking dominant. Obviously tracking scenarios in Kovaak’s are effective and are able to isolate certain modalities of tracking aim such as smoothness or reactivity, and these modalities are thus carried over into the game (like Apex) and by becoming competent at aiming tasks in Kovaak’s, there is a corresponding competency in aiming mechanics in game.
So where exactly does this fail when trying to replicate the results for tacfps? Looking through most tacfps routines, it appears that a solid majority of the scenarios being played are static scenarios (such as 1w4ts or 5 sphere hipfire), as an attempt to replicate the low ttk present in tacfps games. This isn’t really surprising. However, static scenarios have a different set of modalities compared to tracking. The two primary modalities in static are speed and smoothness.
For my purposes I will define the speed as the speed at which the crosshair traverses as one aims at the target, and the smoothness being the efficiency of the people pathing to the target. The essential problem in training for TacFPS is that the training fails to replicate these modalities between the aiming tasks in Kovaak’s and in-game.
For example, when playing a static scenario, ideally you aim for greater than 90% accuracy in order to maximize your score on a given run. However, in game, we see that statistically the accuracy is nowhere near that. In game your target is almost always the head, and yet we see that professionals often have sub 30% headshot percentages. This isn’t even to mention the fact that the actual first shot accuracy is likely way lower than even that as 30% is only considering all the shots that actually hit the opponent.
With such a huge difference in the accuracy, there will be a correspondingly large difference in the speed required for tacfps vs Kovaak’s. The reason for this difference is obvious; in most tacfps duels, you are forced to kill your opponent as fast as possible, or else they kill you first. This factor is not present in a static scenario.
It’s important to note that this difference isn’t simply because players don’t understand the “optimal” way to aim, or any of that “slow is smooth smooth is fast” crap. In fact, I believe that the way in which you aim in a game, is likely close to optimal for your abilities.
Regardless, if you really need any proof, the following thought experiment will suffice: Say you have two equally skilled opposing players, and all their mechanics are equalized in some optimal scenario. They have a pure aim duel in which the difficulty of the shot they have to hit is exactly the same. For simplicity’s sake, say each player has only one bullet, and their intent is to win as many of the duels as possible. Of course, the higher the accuracy goal, the lower the speed. The optimal first shot accuracy in this case will be 50% for each player, as if they aim for a higher first shot accuracy such as 90%, their opponent could undercut them and aim for 80% and would win 80% of the duels as they would’ve fired the first shot every time and hit it 80% of the time. This continues until they approach optimal at 50% whereby any deviation from this accuracy is disadvantageous (this is known as a Nash Equilibrium in Game Theory).
Now that we have that out of the way, that brings us to our problem. Static scenarios (the commonly agreed upon premier scenarios for tacfps) fail to emphasize the modalities of aim that are important when actually in game, causing it to lose a bunch of specificity for game-specific aim and mouse control. While static scenarios tend to favor a lot more accuracy and thus smoothness for efficient pathing, it loses a lot of speed compared to how one aims in game.
If you have a half a brain, you’ll ask, why not use static scenarios with larger targets to emphasize speed? Now this is actually a good question, but here’s the kicker. The problem is that these scenarios do not require precision.
What does that mean? Say for example you are moving your crosshair at a given speed towards a given target. The range of error for where you ultimately land your crosshair can be given by a function that relates error distance assumed to be uniformly around a point in a circle (this assumption is wrong but I’ll get to that later) at the center of your target. Obviously, as distance increases, the probability of landing your crosshair at that distance is less and less. So the point is, there are no static scenarios where you are forced to combine both speed and precision to the same extent required in game. On the one hand you have large targets where you have to utilize a lot of speed, but the targets are so large that your margin of error is incredibly high, and on the other hand you have small targets which, while having a small margin of error, fail to force you to utilize a lot of speed. In our case, we need to aim at a small target (a head) at a fast speed (<30% accuracy).
Now let’s try to understand this at a more fundamental level. When you aim at a target, and are attempting to acquire it in your crosshair (this is how most click timing/target switching is), you attempt to move your crosshair to the target and keep it there. In other words, there are two components, the pathing your crosshair takes towards the target and your ability to stop on the target. Now, you’ll notice that these two components are, in fact, very similar to the modalities of aim that we discussed previously - smoothness and speed. The similarity between efficient pathing and smoothness is obvious of course, efficient pathing suggests that your crosshair moves directly in the direction of where you intend to aim with as little deviation as possible.
However, the commonalities between speed and your stopping power is more subtle, until you realize that speed is limited by your ability to stop your crosshair precisely more than your ability to efficiently path at that speed. This brings us back to the point where I likened the range of error of where your crosshair will end up compared to where you’re aiming, as previously we were working under the assumption that this range of error is roughly circular as in at a certain x error distance from target, there is one corresponding probability of landing one’s crosshair there. However, in fact, this is certainly not the case.
The reality is that the range of error is more akin to an ellipse, whereby a direct line from where the crosshair originally is to the target passes through an ellipse which has a center on the target (So if your target is directly horizontal to your crosshair, you’re more likely to miss to the left and to the right of the target than above or below it). This is because the primary limiting factor in landing the crosshair precisely is stopping power. The ability to path efficiently towards a target is not affected nearly as much by speed as the ability to stop the crosshair on the target.
We can observe this empirically and you can test this for yourself. Go into a static scenario and attempt to flick to a target as fast as possible. You’ll notice that for the most part you’ll “overflick” or “underflick,” hence why those are actual terms that are used. Assuming that this is true, theoretically the elongation of the error ellipse should be directly related to the speed at which you’re aiming. As you aim slower and slower, you require less stopping power, and thus are less likely to overflick or underflick.
Now we’re getting to the good stuff. In my opinion, this is what differentiates a lot of the aiming in tacfps and Kovaak, or other Aim Trainers. The problem is that most static scenarios don’t require much stopping power at all. In fact, it’s optimal if your error region is as close to circular as possible, because that maximizes efficiency. Therefore, you’ll end up having smoother, less abrupt deceleration of the mouse, simply because it’s the optimal way to aim in aim trainers. In tacfps games you don’t have this luxury. They emphasize speed and precision, not accuracy.
This is also compounded by the fact that most of the aiming done is along the horizontal not vertical axis. Since horizontal movement is typically performed along 2-3 hinges (arm, wrist, and possibly fingers), one more than vertical movement which (arm and possibly fingers) and furthermore vertical movement is dictated by the shoulder joint in congruity with the elbow joint, which results in a longer, and presumably less precise lever, one’s stopping power is greater in the horizontal rather than vertical axis.
For these reasons, I believe that most of the scenarios people play in Kovaak or other aim trainers, are best at improving the efficiency of the pathing towards a target, as in controlling their mouse while moving it at a given speed, rather than stopping power, where one must be proficient at accelerating and decelerating their mouse in a rapid and controlled fashion. What needs to occur therefore, is some form of training that better helps isolate this part of mouse control, which is why people experience hugely diminishing returns on the impact of traditional aim training in Kovaak’s because they fail to do so. So with all that theory out of the way, let’s get to the good stuff: the routine.
The Routine:
I am going to preface this with the following disclaimer: This routine will only really work to its full potential if you explicitly play the scenarios in the way I outline. Otherwise it will simply be yet another tacfps routine that is marginally more or less effective than what’s out there. The instructions in this instance are far more important than the scenarios. Additionally, unlike many other routines, the intent of this routine is to help practice the mouse movement and mechanics needed in game, rather than mimic in game scenarios or aiming, as I believe this is better trained using in-game resources, hence why this routine must be supplemented with deathmatch or whatever equivalent there is in your game of choice. A good resource is How To Effectively DM In Valorant by Prodigies.
Routine 1:
(Recommended sensitivity is in-game sensitivity, assuming that the user of this routine is within the range of 35-70 cm/360) This routine is intended to focus on the raw stopping power and speed aspect of aiming, which ideally helps speed up the initial flicks towards the target and how rapidly one is able to acquire the target in a precise manner. For all of these scenarios you should be holding down mouse1(left clock or whatever is binded to shoot.) You should be focusing on “snapping” onto the target as fast as you physically can. It’s okay to miss your flick by a lot, but ensure that you attempt to acquire the target as quickly as possible. This means your flick should be fast, and your adjustment should also be as fast as possible, no need for one clean adjustment. If you start shaking like you have Parkinson's this means you are going fast. Smoothness is completely irrelevant here and is actually discouraged. The goal here is to gain mastery over this type of aiming. This routine will combine both target switching and static, in order to best mimic the mouse movement involved with the rapid target acquisition that we are trying to focus on.
voxTS Voltaic mini- 7 runs
This is your run of the mill TS scenario, mostly in the horizontal plane, combining both short, medium, and long distance flicks between targets.
DevTS Goated NR Static Small - 5 runs
This scenario spawns static TS targets over a smaller FOV, which better emphasizes the smaller flicks that are more common in tacfps, assuming you don’t have doodoo crosshair placement and awareness. This is the most common type of duel, and since the range of motion is smaller, it’s often harder to accelerate and decelerate your mouse rapidly over this range, so it needs specific emphasis.
xenTargetSwitch - 3 runs
Here we have more verticality than either of the previous scenarios, just in case, we need a little more agency over that range of motion, especially important in Valorant where broken agent mobility abilities mandate that you have this skill.
Pokeball Frenzy Auto 1w4ts - 5 runs
Pokeball Frenzy Auto Wide Wall 3 Targets - 5 runs
Pokeball 5 sphere hipfire extra small - 5 runs
These last few pokeball scenarios are essentially following the same logic of the TS scenarios, but now we incorporate static scenarios because they are smaller targets and also you have to stabilize your acquisition on target
Flicker Plaza final bot - 5 runs
Just play this only focused on reacquiring and snapping back onto the target as fast as possible when it “blinks.” This kind of mouse movement/aiming should kind of help simulate the sort of microadjustment and “locking” onto a target especially when they swing an angle and your crosshair placement is off.
Routine 2:
(Recommended sensitivity is between 25-35cm/360) This routine is intended to help bridge the gap between the previous all Speed Routine and bring more and more smoothness and consistency to that speed.
fuglaaPressure 6 bots - 5 runs
This scenario is intended to get you used to aiming faster than you’re comfortable with static targets. Make sure you try to hit every target before it pops and goes away. Don’t worry about missing, the intention should be to at least attempt to hit every target if possible. If too difficult (<60% acc) you can try the version with 4 targets.
1w4ts Voltaic - 10 runs (7,3)
6 sphere hipfire voltaic extra small - 7 runs (5,2)
Wide Wall 3 Targets - 5 runs (3,2)
For these 3 static scenarios, observe the numbers in parentheses. The first number represents how many “speed” runs you will do. During these runs, your only goal is to push up your “shots fired” count. Keep pushing speed, but don’t let your accuracy slip below 60%, but you should be going at a speed where your accuracy is below 80%. You’ll know you’re doing it right when you notice that you are starting to use some of the “stopping power” that you trained in the previous routine with your initial flick and as a result tense a tiny bit. The second number represents the number of “accuracy” runs which you will do after the speed runs. For these attempts to get at least 90-95% accuracy, it will require a lot of discipline to make sure you aren’t going too fast, after pushing your speed for the previous runs. You will likely more or less maintain speed on the initial flicks, and then focus on making sure your micro adjustments land on target.
Reflex Micro Flick 250ms - 5 runs
Reflex Flick - Hard - 3 runs
These last two scenarios should be relatively difficult. Try to hit the target with one motion to the target and back to the center. If too easy (>50% accuracy) I would recommend time-scaling it faster. If too hard (<25% accuracy) I would recommend doing the opposite for the Micro Flick scenario, for the other you can try the easier versions.
Routine 3:
(Recommended sensitivity is between 20-25cm/360) This final routine is intended to focus solely on smoothness/pathing, working the opposite end of the spectrum from the first routine.
1w4ts reload 30% smaller - 10 runs
10 sphere hipfire extra small - 5 runs
Wide Wall 4 Targets small - 5 runs
Pokeball Frenzy Auto Small Wide - 5 runs
Play these 4 static/pokeball scenarios with the intention of keeping your pathing towards the target as efficient as possible. This means straight as possible lines, as little micro adjustments as possible, with clean flicks that stop right on target. You will have to move your crosshair pretty slowly to achieve this.
Smoothbot Invincible Goated - 5 runs
Centering II 180 no strafes - 5 runs
These two smooth tracking scenarios are also intended to help smooth out your pathing. They force you to be able to track a slow moving target precisely without shaking off, which should help with cleaning up the flicks and micro adjustments.
The way I would recommend using these 3 routines would be in succession, doing one a day. Personally, it is my opinion that routine 1 should be repeated with the highest frequency for the average person, as I believe what is practiced in that routine is the most dominant part of aiming in tacfps. However, as a whole all 3 routines are intended to provide a holistic approach towards developing your aim without any substantial weaknesses.You can also choose to repeat routine 1 or 3 with higher frequency if you believe that aspect of your aim needs more work. For example if you wanted to focus more on your speed/target acquisition you could play routine 1 twice, then follow it up with routine 2 and 3, and cycle through that. If you’re unsure I’d just stick with the default one of each, as it should be more than effective enough.
Here are the routines in order. I highly recommend you read this all first (Also do others the favor of recommending they read the doc before doing the routines so they don’t miss the instructions on how to play the routine.)
Kovaak’s Routines Download:
https://www.mediafire.com/folder/bchng96mewqzn/Aimgud_Routines
Aimlabs Routines:
https://steamcommunity.com/sharedfiles/filedetails/?id=2568089866 https://steamcommunity.com/sharedfiles/filedetails/?id=2568089996 https://steamcommunity.com/sharedfiles/filedetails/?id=2568090118
TLDR: You’re doing AimTrainer’s wrong, here’s how to do it correctly :)