Aight, so nobody reads long suggestions, so I’ve done my best to make this one short.
Unfortunately, most people on this forum probably can’t be asked to read that either, so I’ve made lots of pictures to hopefully capture your 20 second attention spans.
If even the pictures are too much for you, you can scroll to the bottom for a tl;dr.
To start with, I’m going to assume you know that:
- The percentage stats (agility, magic size, casting speed, etc) are garbage at our levels.
- The problem won’t solve itself as we level up, regardless of what the formula is.
If you disagree with either of these statements, you are objectively wrong.
Click here to find out why!
Why doesn’t the current scaling work?
This is a tricky thing to explain briefly because we still don’t know exactly what the current scaling is. This is why I put off making this suggestion so long.
However, we know enough to know that it won’t work, and that it has WoM sized problems that can only be fixed with an AO style approach.
The tl;dr is:
- It’s either going to:
- Be too big at high levels.
- Make mixed builds broken at high levels.
- There’s no sweet spot in the middle, just the worst of both worlds.
- It will always be confusing, and useless at low levels.
.
For a "slightly" longer explanation:
So the current formula defies analysis, but at our levels it does this:
If you look closely, you’ll see that trendline is slightly curved. That’s because as you use more agility, you get slightly less bonus per point. The curve is much more obvious if you plot percentage against agility per point of agility, but I won’t bore you with the maths.
unless you're into that kinda thing
look at those curves 
So what does it do at high levels? One of these things:
It either continues doing what it looks like it’s doing, and goes relatively linear all the way up to 1000% or so.
Or it suddenly (and unexpectedly) curves off rapidly, similar to meta’s original suggestions.
The first one is obviously bad (nuking entire islands with self explosion is only funny the first time), but what about the second one?
That’s also bad. Let me explain why:
The idea behind this system is that if I’m at level 100 and go full agility, I can get around +50%. If I’m at level 200, I can get around +100%. If I’m level 1000, I can get +150%. The value always goes up, gets to a good value quickly(ish), and doesn’t get too high.
Seems reasonable, right? It was reasonable enough for 140 people in June.
The problems are more sneaky:
- If I’m level 1000, and put on the level 1000 equivalent of a minotaur helmet, I’d get +100% in two stats, only using one slot and no enchantment. This is very OP, and means that investing more than one percentage stat item into a build is basically pointless.
- Everyone just goes “two items to get +100% in every useful stat, and the rest of the build is power/defense”.
- Past level 200 or so, there’s almost no point upgrading your gear for anything that’s not power or defense. The difference between a level 200 and level 400 agility helmet is almost too small to notice.
How come power and defense work?
This is a quick one - power and defense have one property that make their scaling work:
- The player’s base stats scale linearly with level.
- This means that level 100 gear at level 100 will give you (approximately) the same boost that level 200 gear at level 200 does.
- The scaling is pretty fast. Most of your stats are base stats from levels.
- This means that you always need to upgrade your gear. Level 100 gear doesn’t work at level 200.
So we can just make agility and magic size and all the rest work the same way, and problem solved, right?
Unfortunately, no. While it’s ok for a level 1000 player to have 50x the damage of a level 1, it’s not ok for a level 1000 player to have 50x the magic size.
What we need is a system that:
- Has a significant improvement at every level (so you need to upgrade your gear, and mixed builds don’t break the game).
- Starts at a good value (so stuff’s viable at low levels).
- Doesn’t get too high (so stuff’s not OP at high levels).
- Ideally, is simple and easy to understand.
When you put it like that, it’s obvious why the current approach isn’t working.
So how do we fix it?
So if we can’t scale the percentage stats without them being useless at low levels, overpowered at high levels, or pointless to upgrade at high levels (or a combination of all three), what can we do?
The answer is simple: Don’t scale them.
Put a small amount of power, defense or strength on every item, and scale that instead.
This has many big advantages:
- The most obvious one is that percentage stat gear becomes good at low levels.
- Previously, a max level bandana and hood would be +6.5% and +10% respectively, and get worse if you use more agility items. Since we don’t need to worry about future items being OP, we can make low level items provide noticable improvements.
- Another big improvement is that it’s clear what the items do now.
- Science still can’t answer “what’s +5 agility actually do?”
+5% agility is far clearer, and helps players make informed choices about their gear.
- Science still can’t answer “what’s +5 agility actually do?”
- Different items for the same stat can now be unique.
- Currently, there’s literally no difference between Crow Wings, Dual Armbands and the Fur Hood. With this suggestion, they could all have different secondary stats, making them all unique!
What about enchantments?
Obviously, powerful, strong and hard are unchanged.
What about amulets?
There’s no obvious secondary stat to give amulets, and if an item has no secondary stat, it can’t level up.
I don’t want to change how amulets work for the linear stats, so I’ve instead replaced amulets with charms.
FAQ
Doesn't this make percentage items OP at low levels?
Only at level 10. Power, defence and all the rest catch up by level 20, and everything is roughly the same past that point.
If you don’t believe me, look at this graph of maximum percentage improvement against level:
Power is very weak at level 10 (because it’s enchantment doesn’t do anything yet), but aside from that everything’s pretty similar.
Agility and Magic Size look high on the graph, though remember that these are maxed “glass cannon” sets (and agility is half effective while sprinting). They spike at level 50 because they unlock their first armor set, and don’t increase that fast for future tiers.
In practice, you won’t see values that high in a viable build.
In this thread I’ve leaned towards over-buffing the percentage items. I don’t think a percentage stat meta is a bad thing - there’s 5 of them and it’s far more interesting than the power/defense meta.
All these numbers can be rebalanced if it turns out this stuff is OP, but the main point of this suggestion is the new scaling, not the exact numbers.
Doesn't this obsolete power and defense gear?
Nah.
Remember, a level 50 defense accessory is providing +6 defense per 10 levels, or higher for rare items.
A level 50 agility accessory providing +2 doesn’t really compare. If you want defence, you’ll always pick a defence item. Same for power.
The secondary stats on percentage items are just that - secondary. It’s a small bonus to let you build more heavily into percentage stats, without crippling your damage or health.
Changing every item in the game? Isn't that unreasonable?
Kinda, but not as much as you might think.
This actually only affects 46 items (out of 81), and Vetex has to change every item in the game anyway for the new level caps. Since culture items have just been removed, and new items haven’t been added yet, there will never be a better chance to do a big change like this.
