I feel like my main disappointment with the class trees is some seemed to give little consideration to making talents proportional to their level. Or, alternately, it was decided some talents would “pay off” earlier, but since you get 3 to pick from that adds a ton of variance to the power of each class.
For example, I’m leveling Thief right now. I find Level 5 to be somewhat “balanced”. Let’s keep in mind that to reach level 5 requires 15 XP, or winning 7-15 battles. It lets you choose between:
- Gain 2 magic if using a Dagger.
- All Naga allies gain 1 Armor every turn.
- Gain 2 magic if using a Missile.
These are all small and conditional perks. While “gain 1 armor every turn” feels more valuable than getting a static +2 magic, it only applies if the hero is with a Naga team, so all of these choices are roughly as strong. A “problem” is we can imagine many teams where none of these perks apply, but that’s OK! It only cost us about 10 minutes to unlock this so it seems reasonable that it’s not stunning.
Now let’s turn our attention to level 100, which requires more than 5,000 champion XP or between 2500 and 5000 battles to achieve:
- All Rogues gain 1 magic on a 4 or 5 match.
- 7% chance to assassinate the last enemy when another dies.
- All Stryx allies gain 1 Attack every turn.
What. The. Heck? “Do this on a 4 or 5 match” is neat, but we gained one of those way back at level 1 that required 1 combat. “All {race} gain 1 {stat} every turn” was available at level 5. So the only talent here that feels worth playing literally thousands of matches is the unconditional chance to kill an enemy. The other two talents will literally do nothing in most team compositions, and aren’t very powerful in their narrow best case!
But I think this happened because all three of the trees have comparatively powerful abilities at 70 (~2500 XP):
- 30% chance to dodge skull damage.
- Poison a random enemy when matching Purple gems.
- 25% chance to summon a Heronath when an ally casts a spell.
While there’s one I think players consider the strongest here, these are all fairly powerful skills in a broad context. The summon is the “weakest”, but considering some other talents in the tree there’s evidence Thief can be a very summon-heavy build if that is desirable.
What I’m getting at: in 2 of the 3 skill trees, level 70 has a better skill than level 100. 70 is half the effort of 100. I think the story for those trees is “It got a comparatively powerful level 70 skill because its level 100 skill is lackluster.”
But that leaves our Thief class with only one viable level 100 talent. Worse, it opens the door to a class that gets 3 trees topping out at level 70, meaning it gets a garbage level 100.
Wrapping up…
I think players would be much happier if every tree’s skills at level n were at least arguably comparable to each other. If we’re going to put “gain 1 stat point at each turn” at level 5 for some tree, it’s somewhat ludicrous to put it at level 100 on another. No class should have 2 lackluster talents at 100, considering it takes twice as much effort to get there as level 70. I think level 100 talents should be roughly equivalent to legendary/mythic third traits.
The imbalance, particularly at 70/100, is a big part of why the Lightning Strike classes are so powerful. They don’t have to give up anything to take this powerful trait, and there’s almost no comparison even at level 100 amongst other classes. If we’re to have “interesting choices”, there should be some sacrifice involved. Imagine if Rock Solid were a level 70 talent. Which is better, “barrier on brown” or “explode a random gem on 4/5 matches”? That’s a hard question. We never have to ask it, because nobody ever has to choose between those two exceptionally powerful traits with dramatically different costs.
It’d probably be much smarter if each tree had a “theme” per level, something like:
- 1 (1 battle): “The hero counts as {race}.”
- 5 (3-5 battles): “Gain {bonus} if using {weapon}.”
- 10 (~25-55 battles): “Something happens {per turn | at start} for {race}.”
- 20: (~100-210 battles): “Something happens on { 4/5 | color } match.”
- 40: (~400-810 battles): <Equivalent to a 1st trait.>
- 70: (~1250-2500 battles): <Equivalent to a 2nd trait.>
- 100: (~2500-5000 battles): <Unique 3rd trait.>
That would make for more flexibility and actual tradeoffs between abilities. It gives out weak but relevant bonuses for small amounts of leveling effort. And it doles out powerful abilities at the final stages, where each level requires roughly twice as much effort as the last.