In terms of build flexibility, you could reduce the number of skill levels, release more armor pieces with this skill, introduce armor pieces with Level 3 of this skill, or make this skill available through Driftmelting.
In terms of reload chance, you could increase the success rate, or add a pity mechanism (e.g., failing at most once).
Current Status
It is a 5-level skill with only 4 armor pieces carrying it, and it cannot be Driftmelted. If the weapon itself does not have this skill, you must use 3 armor pieces to max it out, resulting in low build flexibility.
Taking Levels 4 and 5, where the reload chance is at its highest at 70%, as an example: in actual combat, in the worst-case scenario it can take 4–5 dodges to successfully reload. This situation occurs more than once in a single hunt, and the success rate feels far lower than 70%.
If you’re going to talk about probabilities, shouldn’t you provide evidence?
From your wording, it just looks like you’re criticizing based solely on impressions.
Reloading while moving with a 70% chance, ignoring Reload Speed
not switching to the next ammo type
is an exceptional skill, I think.
Depending on the ammo type and situation,
if you consider that you get an advantage equivalent to a Perfect Evade 70% of the time,
I think you can see how strong it is.
For game balance,
having few armor pieces with it, not being available as a Driftsmelt skill, and not reloading every time
seems reasonable to me.
I don’t think “Light Bowgun’s Evasive Reload” needs any adjustment, as it already makes it possible to keep firing powerful ammo types just by swiping.
However, Gunlance and Charge Blade, which are also affected by the skill besides Bowguns, rarely get a chance to use it, so I think adjustments such as individually increasing the probability for those weapons are necessary.
I think chance-based skills are meant to be used while accepting that RNG can be streaky.
It would be a different story if there were a monster you absolutely couldn’t beat without using Evasive Reload, but
I think it’s on the same level as saying,
“I lost because Brutal Strike didn’t high-roll.”
That’s what you’d call cognitive bias.
Using a chance-based skill means accepting the risk of bad RNG.
You can calculate the expected value, but that doesn’t guarantee consistency when actually using it.
Anyway, this is getting off-topic, so I’ll leave it at that.
Even though I usually dislike Evading Reload gunners, I really think they should add more pieces where you can fit Lv1+α.
The head is the only piece with Lv1, and the extra skill doesn’t directly contribute to damage, so compared to how easy Lv4 is to fit, the cost-performance of running Lv5 is honestly hopelessly bad.
If it were something like Evading Reload Lv1 + Recoil Down Lv1, it wouldn’t benefit Doshu or crouching HBG as much, and I think it’d be good since it would open up more options for other bowguns.
Failing to reload consecutively probably means that PRNG compensation (pseudo-random chance) hasn’t been implemented.
*PRNG compensation = A system where the probability increases after consecutive failures
Pros of implementing PRNG compensation:
Consecutive reload failures would virtually disappear
Cons:
Increased server load
Increased data traffic
Increased risk of unexpected bugs occurring
Increased development costs
Disrupts game balance
As for that last point about disrupting game balance, I think you’ll understand if you look at Brutal Strike as an example.
Well, I’ve written all this, but keep in mind it’s all just speculation.