Building a board of different weapons, there needed to be reasons why the weapons would be useful in different situations and you'd not just always take whichever does the most damage. Maybe some do less damage but build up resources so you use those against weaker enemies to prepare for stronger ones. Maybe some do more damage against specific enemy types, maybe like a system of elemental resistances and vulnerabilities? This is a FIRE weapon it does 25% more damage to ICE enemies but 25% less to STEEL enemies blah blah blah.
I'd been playing Magic with the recent-ish set that used Poison/Infect and finding a lot of joy in the moments of choosing how to block when some attackers are poisonous and others do regular damage and you can't stop them all. Just quite nice to be choosing between different kinds of damage, having to estimate which one will come closest to killing me, much more pleasant than the standard of "the biggest number is worst". They're good moments, maybe I can steal the system of damage types and put it in a new context. Life and Poison function differently - Life is lost / Poison is gained - but that's unnecessary, losing is just winning backwards.
(tools: just keep stealing from Magic it has more stuff than it needs anyway.)
Also talking with Jonathan Brodsky about Nightmare Cooperative and trying to articulate what makes roguelikes work, that thing of being presented with random situations and having to choose which resource to spend to deal with them. Just having random situations isn't enough on its own, you need an option to bypass some of your choice. And you want that choice to be unclear else it's a pure optimisation puzzle, acting under uncertainty being part of the pleasure of strategy games (in principle you could work out all the probabilities and pick the one with the highest survival chance but in practice you can make it sufficiently complex that some estimation is required). Usually the choice is broadly between using the basic combat mechanics (spending some number of hit points) or invoking a limited-use secondary system like items or spells.
Having different damage types would put more of that resource decision into the basic combat mechanics. Flanked by two enemies, the choice isn't just "use an item or take damage", it's which one you want to deal with first while taking damage from the other. Conventionally this is a simple calculation, one value to optimise, kill whichever one would do the most damage. But with two hit point values, you can't locally optimise, you have to estimate how much of each you'll be taking afterwards too.
I think this is a really weird and technical idea! It's already complex in Magic but I've extended it to enemies that move about on a grid, and to everyone having the two hit point meters (not just the player) because half the point is that your weapons can deal two kinds of damage too to give reasons to use different ones against different enemies. Wow so complicated, double the numbers everywhere. Then I start to implement it and realise it's just the same old red and blue life bars from every single CRPG ever. Oh !
Yeah so actually it is nothing special! I thought I was being clever in how I took pieces from different games and recontextualised them to make something new, instead I just made something old. Basically every game ever has blood and mana as separate resources. Usually the second one you are spending yourself rather than getting hit in but sometimes there are energy draining undead monsters hitting it so all I'm really proposing is to have more of those. So conventional game. (Also I may as well just make some weapons that spend mana too because it'll be plenty intuitive, everyone expects it.)
Anyway! To make it an interesting choice which kind damage to take I want to make the things that threaten them be actually different. Not just red and blue flavours of the same enemy. Enemy design is probably a separate post but here it suffices to say that giving them different strengths and weaknesses will probably complicate that enough, make it non-trivial to estimate which ones your board is best able to handle in the immediate future. Spending mana for positive effects also adds a nice twist; if you lose it in damage you can't afford the bonuses.
I've noticed that players very quickly grasp the idea that enemies are weaker in one colour so it is advantageous to hit with weapons of that colour, but then often they internalise that as a strict rule ("you always have to hit those ones with a blue weapon") and will go out of their way to use the preferred colour even when it's very unsafe to do so / costs a lot more turns. There's definitely situations where it's better to use the "wrong" colour.
Also, different amounts of damage of each colour, whole matrix of possibilities. I had a weapon that did one of each but dropped it, important for everything to have weaknesses. I really like the way having weapons do different amounts of damage can create "combos" already before any extra rules come into play: an enemy has 3 hit points, you can hit it once for 2 damage and then finish it off with a 1-damage weapon, combo! Much more elegant than the labelled combos of "apply a status effect to an enemy" / "trigger off enemies with that status effect" though it's nice to have those too.