A while back, I wrote these wand rules, which have been active in my Vaults of Pahvelorn OD&D game. I still like them just fine, but they are a bit more complicated than needed, and several of the flourishes (such as the final strike), have never actually been used, and could probably be removed without much loss. Below is another, simpler system for wands in traditional fantasy games that I think might be an improvement. It preserves the same basic dynamic of allowing magic-users to target an enemy’s save versus magic (which is a proxy for “magic defense”) rather than armor class.
The power of a wand is measured by damage die size, and follows the progression of d4, d6, d8, d10, d12. Wands cost 100 GP per die size and can be replenished or improved as a downtime action for 100 GP per die increased. All wands inflict damage of a specific elemental type, which is determined upon creation, and elemental damage may have additional effects depending on circumstance. Common elements include fire, ice, and lightning.
Wands may be used to attack a single enemy in sight within 60 feet. Damage inflicted is determined by rolling the wand die. The target then makes a saving throw, and decreases damage taken by the margin of success. If the wand die comes up 1, the wand die size decreases. If the wand was already at d4, it becomes exhausted until replenished.
For example, a magic-user buys a wand of fire d8 for 300 GP. This wand may be used to make any number of ranged fire attacks until the damage die comes up 1, at which time it becomes a wand of fire d6. The wand does fire type damage and thus may also ignite flammable materials, melt frozen objects, and so forth.
As an out of turn reaction, no more than once per turn, the wielder of a wand may counter another wand attack. Though no damage is dealt in either case, both wand dice must still be rolled to test for wand exhaustion.
This method does give the magic-user more combat options (though without completely avoiding resource management), and as such may not be appropriate for all campaigns. It is possible that the costs might need to be adjusted, but this will likely depend at least partly on other elements of a particular campaign setting. In the past, I think I have erred in setting the cost of consumable items too high to be attractive to players. Consumables need to be relatively cheap to seem worth it. Just expensive enough to seem like an actual expense, but not much more. In comparison, the prices in the Expert rulebook seem kind of out there: who would spend 10000 GP for 20 arrows +1 (page X52), even at high level, especially if magic items are sometimes found during adventures or if a party magic-user has access to a renewable attack spell?
Six sided die progression variation
1d6-1, 1d6, 1d6+1, 1d6+2, 1d6+3. These dice expressions have the same expected values as the polyhedral version given above, but smaller deviations. This method is thus slightly more consistent, but also has the potential for zero damage in the first case. That could be removed by specifying that minimum damage is one, at the cost of slightly increasing the statistical complexity (expected damage becomes 2.67 with deviation 1.49 compared to 2.5/1.71). The d6 chain also has a lower upper bound.
Optional capacity below d4
There are states less than d4, which are d3, d2, d1. These all count as a single step together for purposes of replenishment, however, so 100 GP is sufficient to bring a wand back to d4 if it is at any of those lower states. This variation would be appropriate for those that desire wands to be always at least a little bit useful.