This is a *statistical* problem.

Here’s the problem: we just don’t need Majors that much.

A list of requirements for traits on various rarities I posted here gives us a rough idea of just how many of each type we need. Adding up what we’d need to max traits for one troop of each rarity - so all three traits on a Common, Rare, Ultra-Rare, Epic and Legendary - brought me to this total:

- Minors - 294
- Majors - 66
- Runics - 64
- Arcanes - 34
- Celestials - 3

If both the drop rate and the usage rate are a) proportional based on rarity/value, and b) aligned with each other, we should expect to see somewhere between 100 and 150 Majors needed here.

Of course, that isn’t representative of everything, because we have more Rare and Epic troops than the other three rarities. How about some *even more frightening* numbers? I added up the total numbers of each traitstone tier we need to max every single current troop - 25 Commons, 44 Rares, 24 Ultra-Rares, 40 Epics, 25 Legendaries. We get this:

- Minors - 9192
- Majors - 2176
- Runics - 1808
- Arcanes - 1065
- Celestials - 90

THESE TOTALS HAVE BEEN EDITED, because I counted wrong. But its still silly.

(Assuming I’m right about how many Celestials an Epic needs but that isn’t the point!)

As for *getting* the things…

We get proportionally more Major Traitstones than others because the majority of our keys are Glory Keys, and traitstones in those default to Major or higher (and the large chance of Minor Traitstones is locked out, found in Gold Keys). That much, you all already knew.

Now, if we *ignore* that and try to use some more neutral sources … we basically have a standard pyramid scheme.

The drop rate (in battle as well as anywhere else) of various tiers of traitstone is proportionally linked to their rarity/value. A run of 54 battles I just had, all in the same kingdom, gave me the following traitstones: 22 Minor, 9 Major, 6 Runic and 1 Arcane. While admittedly not a particularly large sample size (and I’m fairly sure I got lucky with Runics), it confirms what we know.

So there’s all that. Clearly the devs need to tweak some numbers here.

I’ve discovered my math here is slightly off and edited it. But *slightly* doesn’t explain *that* and the new (hopefully real this time) numbers still suggest issues.