Read on Twitter
A feature that would be really cool to see in an aim trainer would be scoring multipliers conditioned on the difficulty of a given run. This isn& #39;t referring to the general difficulty of a scenario, but the particular difficulty of a run caused by variance in RNG.
A simple example of this could be target distance in a static targets scenario such as 1wall 6targets small. A lucky player might get many clusters of targets that can be quickly completed, while an unlucky player might be dealt consistently spread out targets.
How you should weigh different distances is a design decision, but a good starting point could be calculating index of difficulty using Fitts& #39;s Law ( https://en.wikipedia.org/wiki/Fitts%27s_law).">https://en.wikipedia.org/wiki/Fitt... We don& #39;t know if Fitts& #39;s Law applies to 3D aiming, but it& #39;s been replicated in many contexts already.
So, a very small distance might be weighed as worth less than a very large distance, mitigating the clustering advantage I just laid out. We can even trust the player on this one instead of needing to rely on something like mean distance, as long as we trust the weighing.
Calculating distance here is actually quite tricky, though. Naively, one might assume that you simply take the position of point A and point B in 3D space and determine distance this way. This could work in scenarios where all targets spawn on one plane, but it breaks otherwise.
Consider this masterpiece of a diagram I made: two targets can spawn in the same x and y positions, centered on a player& #39;s view, but be in dramatically different places on the z axis. Here, the targets would both visually appear in the same spot for the player.
The solution is to measure distance using angle of rotation instead, i.e. how many degrees does a player have to turn to go from point A to point B? This is still computable, but more work than simple distance. Still, it is certainly possible.
What is the benefit of implementing this, though? Well, besides reducing scoring variance, I also think it helps player practice better. Currently, pretty much all switching scenarios (click and track) inherently value the closest target in most settings.
Because of this, the best way to encourage and measure more distant targets is to either increase target spawn areas, or use fewer targets. These solutions work, but have their own drawbacks, such as targets spawning off screen more often, and even higher scoring variance.
Dynamic difficulty scoring allows for more player freedom in navigating targets, and even allows players to play to their strengths if they are gunning for a high score. Routing through targets is useful for upping KovaaK scores right now, even though its in-game utility is low.
This thread is a small slice of the work that would be needed to properly implement this feature, though, and is far from covering many other cases. I would love to see it eventually.
Maybe AimLab can use that VC funding for something other than marketing.
https://abs.twimg.com/emoji/v2/... draggable="false" alt="🤷‍♂️" title="Man shrugging" aria-label="Emoji: Man shrugging">
Maybe AimLab can use that VC funding for something other than marketing.
