This was migrated from the summary of the “Alchemy” mega-idea following its breakup into smaller independent ones. It is not an idea in and of itself, but is rather design notes on the present state of the alchemy system as a whole which may still be useful.

NetHack’s alchemy system is balanced around the assumption that the player will not hold onto nearly every potion in the game, turn them to water, and polypile them into alchemy ingredients; however, this is not hard to do. Should the player do this, they can amass hundreds of hit points worth of full healing potions, or all the gain ability potions required to max out every stat, or get dozens of potions of holy water.

For most potions in the game, diluting for holy water or alchemizing/polypiling for alchemizing is their ultimate fate, and there are enough potions so that the player never really has to agonize over what to do with a potion. One solution to this would be to make more potions than just acid effective for throwing at monsters.

The color alchemy and gem alchemy patches are unsatisfactory, as is any system that ties alchemy recipes to the random appearance of potions. This is because it makes the game choose at the beginning whether powerful potions can be produced out of junk potions/gems, or out of expensive potions/gems, or not at all — all before the player gets a chance to learn how alchemy will work in this game. It has a large impact on overall strategy that the player can’t discover until probably the midgame. At the same time, having hardcoded alchemy rules is unsatisfactory because it’s inflexible and carries a heavy spoiler requirement.

Random alchemy is also odd: dipping the same potions in the same way can yield different potion results.