Linear scaling gave percentile stats effectively no effect early, but the token amounts we have now aren’t good enough to be worth using either - log scaling still failed to make the percentile stats actually useful in the early game.
We’ll get a brief period (no formula, but if I had to guess around level 150?) where getting a significant amount of a stat requires build investment and diminishing returns let you spec heavily into one, but it’s really not going to be long before you can get almost a full build’s worth of stats in two different stats by wearing something like a minotaur helmet.
Build variety will happen, but it’ll be extremely shortlived.
“All builds are mixed, with 90% power/defense”
isn’t any better than
“All builds are 100% power/defense”
The best solution is to just remove scaling and handle these stats the same way they were done in AA (flat % bonuses, all items have power or defense and scale that way). Honestly I might take linear scaling over logarithmic though. At least with linear scaling we will eventually reach the point where the stats work properly, even if it is at level 5,000 in 2030.
Logarithmic scaling is 100% an awful failure.
sorry for cutting your post up, but i think there’s just been a misunderstanding
I agree that it’ll be easier to hit things later on - really easy, way too easy. Logarithmic scaling enchantments 100% makes things worse. My magic size/agility comparison in the original post was just to exemplify how little of an effect the natural magic size is going to have - the level 850 meta is 100% going to be defined by logarithmic scaling.
also minor addendum but I doubt 2.5x is the largest multiplier we see
We get 1.5x at 2.5% of the level cap with a single item - there’s no way the logarithm has that steep a dropoff.
The “2x at level 5,000” thing from the trello was referring to just the nimble enchantment, a low level enchantment that doesn’t even make up a majority of the agility in level 80 builds. Vetex 100% plans on much higher multipliers.