IIWhiteHawkII

IDK, maybe these are the results of game-testing where designers concluded that it is correct decision. But to be honest I cannot understand why designers changed the nature of some weapons from TF|2 in Apex.

I liked the idea of FLATline that this gun has actually a FLAT recoil. I mean mostly horizontal-only. You know, it's kinda unusual and interesting by concept. Same was with Alternator, but this gun now is a bit closer to TF Origins.

Spitfire is useless potato. In TF it has pretty big but predictive recoil pattern and really huge damage. It was nice gun to cover friends/ defend position and for suppression. But you could counter it with SMG due to Spitfire's low fire rate. Pretty fair.

L-Star... I REALLY love and enjoy this gun since TF|2. And thank you Respawn Team for swapping it with Devo, and for deleting stupid reload.

But I cannot understand the last change of this gun. Whether it's new or old recoil pattern - this gun is still pretty bad for precise ADS long-range fire even if you master this gun. It literally didn't change anything, IMO. At the same time, now even with good accuracy - it overheats too quickly and one load usually isn't enough take down the enemy. It is primarily suppressive Close/Medium range LMG that is best to "choke" enemies with aggressive spray and very useful in hip-firing. Making R-pattern more predictable - doesn't change thing much, while now I struggle hard to takedown enemy until I am dead because of overheat.... The change in priorities doesn't feel fair enough.

P.S.- of course it's nothing more than speculation based on personal feelings. I don't have the results of game testing and how much worse it could be if devs didn't change balance. But yeah, I still have questions :D

> But to be honest I cannot understand why designers changed the nature of some weapons from TF|2 in Apex.

TF2 recoil was functionally random. Each bullet fired applied a base amount of pitch and yaw, and put an extra amount randomness in any direction on top of that. This put a very wonky skill ceiling on controlling recoil because it was either easy to control (e.g. "up and to the right") or impossible to control; it was primarily dependent on the tuning of that extra randomness value. Yes, the Flatline had horizontal recoil, but it was also impossible to learn because the bouncing left/right were completely random. We did experiment with a more horizontal recoil for the Flatline in Apex, but when it comes down to it, trying to practice and learn a recoil pattern that involves rapid left/right direction changes isn't fun. It's hard enough to do with a mouse, and damned near impossible on a controller. For more evidence of the difficulty of controlling left/right recoil, look no further than the previous version of the L-STAR, and comments in this thread about the Alternator. The Alternator is 100% predictable in it's left/right, but it's still a pain if you're trying to use it at longer ranges (not recommended).

> Spitfire is useless potato. In TF it has pretty big but predictive recoil pattern and really huge damage.

There was no predictive pattern at all on that one, it just kind of saturates a cone with high damage bullets... it's like a slow, relentless shotgun.

Another major factor in TF2 that limited the skill ceiling was that weapons would basically cease vertical recoil after some time. For every upward degree of recoil that is applied to the weapon, there was a "spring" that was pulling it back towards the center. Eventually the upward recoil and the downward force applied by that spring would equalize, capping the gun out.

The behaviors in TF2 were tuned for a much faster paced game that involved tons of firing at high speed targets (basically, pathfinder everywhere), firing while being a high speed target, and dealing with tons of verticality. Necessarily, there is far less emphasis on recoil control in TF2, but that level and type of tuning isn't appropriate for Apex.

Alt_Mayday

Each bullet fired applied a base amount of pitch and yaw, and put an extra amount randomness in any direction on top of that

But isn't that exactly what's still happening with guns in Apex?

Back in S0-S1 it was possible to laser people at long range with the R-301 if you knew how to control it, but then its recoil randomness was increased in a patch in order to "nerf" it, meaning that it's now very dependant on pure RNG.

Wouldn't it be better to increase the raw, predictable recoil patterns while minimizing randomness in order to tie the effectiveness of a gun to the player's skill instead of some Math.random() function? If someone spends time in the firing range learning spray patterns they should be rewarded.

> But isn't that exactly what's still happening with guns in Apex?

Not really. Instead of a small base pitch/yaw offset with high randomness, Apex has hand tuned recoil patterns where each bullet applies a specific pitch and yaw offset amount from it's part of the pattern, with a small variance. In the editor it looks like a squiggly connect the dots with little boxes around the dots (where the box represents the variance).

> Back in S0-S1 it was possible to laser people at long range with the R-301 if you knew how to control it, but then its recoil randomness was increased in a patch in order to "nerf" it, meaning that it's now very dependant on pure RNG.

"Recoil randomness" is such a broad term and one that is often incorrectly translated to "random recoil". The randomness that you're talking about applies slightly around each point in the overall pattern, but we tune the weapons to have minimal to no overlap between those points, meaning the overall pattern should stay roughly the same. Think of the randomness as deforming the pattern slightly, rather than actually changing it. You can see this in firing range by doing repeated uncontrolled firings. The 301, for example, always kicks initially up, slightly to the right, farther to the right, and finally back to the left. During that whole process the amount of vertical recoil tends to lessen (obviously there's more subtlety to it than that, but you get the idea). The 301 does have greater variance to the end of it's recoil pattern than say, the Flatline, which means it performs better in bursts. The Flatline is more consistent but has a much harder to control pattern. Additionally, the Flatline does not take barrel attachments, so it can never get quite as easy to control as the 301.

> Wouldn't it be better to increase the raw, predictable recoil patterns while minimizing randomness in order to tie the effectiveness of a gun to the player's skill instead of some Math.random() function? If someone spends time in the firing range learning spray patterns they should be rewarded.

That's exactly what we do. However, we're not going to completely remove any randomness, or tune it down to where it's not noticeable... even highly competitive games like CS, or games like RS6:S still have a reasonably high variance. Apex is already far less punishing than even those games, which add a huge amount of randomness for moving while shooting.

Additionally, and as illustrated above, the different guns are tuned to have different strengths and weaknesses. The Havoc is easy to control, but has a wind-up. The 301 is solid at the start, but gets more unwieldy with sustained fire. The Spitfire is wild but extremely consistent. The Devotion actually gets easier to control and more consistent over time. To get good with different weapons potentially requires different skills, instead of varying degrees of the same skill.

Alt_Mayday

Not really. Instead of a small base pitch/yaw offset with high randomness, Apex has hand tuned recoil patterns where each bullet applies a specific pitch and yaw offset amount from it's part of the pattern, with a small variance. In the editor it looks like a squiggly connect the dots with little boxes around the dots (where the box represents the variance).

Yeah that's what I meant with "randomness". And thanks for the detailed explanation, you don't know how much some of us appreciate these insights. Not knowing the exact behind-the-scene mechanics is often a huge problem, because it makes any discussion about these topics riddled with half-truths and speculation. So once again thanks for clarifying it.

Regarding the R-301 - this is what I was talking about:

Increased vertical and horizontal recoil. Slightly increased recoil pattern randomness.

Designer Notes: We’ve made these changes as we’ve seen that the R301 is a bit stronger than intended at long range. We’ve adjusted recoil to balance keeping it effective at medium to close range--which is the original intent--but it should be more difficult for players to auto fire and still laser enemies at long distances.

Many players (me included) were very frustrated and irritated when we saw this change in the patch notes. Firstly because we didn't know exactly what it meant (as you said, "randomness" is a very broad term, and it has a strong connotation with "reducing the skill ceiling by increasing RNG"). And secondly because without the ability to use it at long range reliably/consistently, the R-301 became kind of obsolete compared to the R-99. As a consequence the meta went back to the old R-99/PK/Wingman/Longbow (back in S2).

With your explanation this change now seems more comprehensible in retrospect, so thanks once more. I just wanted to use this opportunity to describe how some balance changes appear from the perspective of the players, who don't know all the details and thought processes behind these decisions.

To get good with different weapons potentially requires different skills, instead of varying degrees of the same skill.

This is exactly why Apex' gunplay is so enjoyable compared to games like CS etc. Most guns feel very unique and distinct from each other. Not just across different categories (shotguns vs SMGs vs pistols etc) but even inside of those categories: PK vs Eva, Sentinel vs Scout, Hemlok vs Flatline etc.

But at the end of the day most players still seem to come back to the same 3-4 out of 23 guns. So why not increase each weapon's skill ceiling by adding more unique mechanics, like the Sentinel's shield battery thingy etc? The Longbow could deal more/less dmg based on how far away from the target you are. This would open up a whole new skill to master: besides moving your crosshair on the enemy your positioning would now also determine the gun's effectivity. Or what about giving the Skullpiercer a tremendous headshot dmg increase, but also a body dmg decrease? This would greatly reward players who are able to hit headshots consistently, without being OP. Or a gun/hop up whose dmg/fire rate/whatever increases if you're low on health. This would also create a completely new set of skills and decision making: "do I risk running around with low health, but dealing tons of damage? is my positioning/movement good enough to make this work out?".

These are just some general ideas that might or might not make more guns than the usual 3-4 suspects viable. And sorry for drifting off topic, there's just so many things you start thinking about after playing this game for over a year lol.

Thanks again for explaining these details and giving some insights. It honestly means a lot for many of us.

Thanks for your feedback!

> But at the end of the day most players still seem to come back to the same 3-4 out of 23 guns.

Agreed that this is how things end up playing out. Ultimately I don't know that it's possible get people using a significantly larger portion of the available weapons; the top-tier players tend to home in on whats the best, and all of the other players end up taking cues from them. We do try to move the meta around each season with different hopups and various balance tweaks of course. It's difficult to push on skills too much, especially when a high skill ceiling is involved. I think the Charge Rifle is a illustrative of that... it was actually fairly trash in low/mid skilled players hands, but a player who "got it" was just completely oppressive. We could look more at "kiss/curse" style mechanics, such as your Skull Piercer suggestion, but the same problem mentioned above tends to rear it's head there too. In the hands of most players the weapon will be ineffective because of their inability to consistently hit headshots, unless the headshot bonus is significant. If it's significant enough to move the needle and get lower skill players using it, it'll be an absolute terror in the hands of a skilled pro. But we'll always keep experimenting and looking for more ways to keep the meta interesting and the gunplay engaging.