In a thread on the RPGsite (see post #15), Barbatruc proposes an interesting method for tracking torches and lanterns. He later mentions being inspired by Intwischa’s method for tracking ammo. Talysman drops in a little later (post #26) to mention that he does something similar with wands. (Which, I’ll note, is very similar to Numenera‘s artifact depletion roll.)
For my own reference, I’m going to archive these methods here briefly:
LIGHT SOURCES: In OD&D, set aside a d6 for each lit torch and a d24 for each lit lantern. At the beginning of each turn roll all the dice set aside: Anything that comes up 1 goes out and gets marked off the character sheet. (This results in torches and lanterns having by-the-book durations on average, but introduces an element of uncertainty and variability. More importantly, it simplifies bookkeeping.)
INTWISCHA’S AMMO: The PC has an “ammo die” of a size determined by the amount of ammunition they’re carrying. They roll this die with each attack roll and if it comes up 1, their die type decreases by one size. If they purchase ammunition or find a stash of it during the adventure, they can increase the die size instead.
ALTERNATIVE AMMO: Have your PCs buy ammo in lots equal to the die size of the system you’re using. (d20 in 3.5, for example.) When you roll a 1 on your attack roll, mark off one lot of ammo. (Trail of Cthulhu uses a similar mechanic in pulp mode, but when you roll the 1 you’re actually clicking on an empty cylinder and automatically miss. I’m ditching the “critical failure” aspect of the mechanic and just using it to track ammo.)
WANDS: Roll percentile dice. On a roll of 1 or 2, the wand has run out of charges. (Note: This system doesn’t work if you want the PCs to have some method of determining exactly how many charges are left in a wand.)
What I’m seeing here is a cluster of techniques that I think can be trivially generalized to cover any form of consumable that are likely to be carried in large quantities for frequent use. I suspect it’s particularly useful if you can incorporate it into a general resolution mechanic (instead of rolling a separate die on every single check).
We have been using Intwischa’s Ammo method for the last few months and it has been working just fine for us.
Other than Intwischa’s method, these seem very harsh.
The negative consequences of something lasting less long than you expected and being unavailable when you need it are much worse than the benefits of something lasting longer than expected, across the board. If your last wand of cure light wounds runs out in the dungeon in the first six uses, that’s a big deal, but if it keeps going for more than 280 uses (which is equally improbable, I ran the numbers), you might notice that you’ve gotten a bit lucky. Adding variance always tilts against the players.
This is less of a problem for ammo, since you expect to be running low and having to refresh it many times in an adventure, probably several times a night. It’s a much bigger problem for torches or wands; in a month-long Underdark expedition, it will probably even out, but in a couple-night dungeon crawl you’re prone to catastrophic runs of bad luck. You can choose the roll or the duration so that things last longer to compensate for the elevated consequences of failure, but if you tell your players the specifics, they’ll probably calculate their needs using the expected value (because humans are bad at probabilities), play things significantly riskier than they think, and end up making themselves *more* likely to get screwed.
With that in mind, I’d suggest one of two things: use dice that form a bell curve with the expected length about 10% longer than the stated fixed duration, or make everything use a very small increment where players will go through a dozen increments each in an average session so that it forms a natural bell curve. This second option is basically a variation of Intwischa’s method that can be better integrated; you can have a two-tier track (i.e. how many magazines, how full is the current magazine), with each failure on the die roll marking down the small track and the small track rolling over onto the large one. This requires a little more bookkeeping, but less annoying than marking every time. For wands, I’d probably go with 1-2 on 1d20 depletes it, a wand depleted 10 times is empty.
I like the idea behind this, but it is just too random in execution.
Running out of a vital resource near the start of a mission is very bad, and pretty unreasonable.
I can’t imagine a veteran soldier would go into battle with an empty magazine or flat batteries, but this will happen not infrequently if you need to roll all the time.
Maybe a better solution would be a 3 strikes thing. This first time you fail the roll you notice it getting low, but don’t have to roll again in the same scene. The second time you notice you are almost out. It is not until the third failure that you suddenly run dry.
The median time to fail for a 1-in-X failure check is ln(0.5)/ln(1-(1/X)). So 13.5 for a 1-in-20, 3.8 for a 1-in-6. Over half will have failed on 14 checks or 4 checks respectively.
My own experience with trying it 1-in-20 for Wands (mystery charges left when found) was that there’s a lot of extra rolls, and in general that your wand always has a lot of charges left but also always fails after just one more roll most often.
So going back to the powerful/strong/moderate/weak/faint ID system (~45/35/22/2 charges), start on a d10, drops you to a d12, which drops you to a d20, which drops you to a d3. But then I’m bookkeeping again, on top of the dice. It’s nice tension-building for players like that though for a really important item like a Staff of Power (not ubiquitous 3e wands). Failing any particular step early is still likely, but failing all of them early is not.
For something like OD&D and Basic, almost all the timekeeping is broken into 6 or 12 or 24 turn lots, so on a 6-wide time sheet (for wandering monster checks) spells and lights run out so many lines below the turn they were started. It’s much easier than rolling dice all the time.
Similarly, arrows can come in multiples of 12, because with 2 or 3 or 4 shots per round you still have nice even numbers to work with. If your magazines hold 20 or 30 rounds because that’s realistic, your game needs to use a 5-round “burst” as the standard abstraction, perhaps with a 10-round “long burst” option. The 6-round revolver uses a different mechanic, same as the double-barrel shotgun.
I first saw this method for Ammo in the 1996s Hong Kong Action Theater!
As much as I like this idea, I’m relatively certain that it won’t be long until the first player asks: “I need to be sure, man – how many bullets have I got left in my mag? Enough to take these six guys?”
Schrödinger’s magazine, if you will… and that brings us back to square one.
Still, it works quite well if everyone at the table is comfortable with maintaining a certain level of abstraction and uncertainty. In my games (3.5) I’m afraid it’s unlikely to last very long.
Of course, your mileage may vary…
My issue isn’t the randomness – its coming from collectible card and miniature games that often use dice as counters. It can just see the PC with a quiver full of arrows some torches and rations. One table bump and he’s suddenly been completely resupplied! 😉
But I do like the idea of generalizing it into a single mechanic. I’m kind of thinking maybe some type of critical failure?
Of course, you don’t roll for torches, but many games have rules for “If you let me screw you, you get an XP.” type of thing.
The key to the mechanic is that you have to be willing to treat it as an abstraction on both sides of the equation: If you’re wedded to the idea that 1 attack roll = 1 bullet, then you’re going to run into the “Schrodinger’s magazine” problem and you should stick to concrete values.
If you’re not wedded to the 1 attack = 1 bullet thing, then the abstraction will hold up: You’ve only got a handful of bullets left. Will that be enough? Guess it depends on how well you shoot, eh?
(Another option would be to include something like a Careful Shot action: You’re only shooting only bullet, so you don’t need to make a reload check. But since you’re being conservative in your shooting, there’s a -2 penalty to your attack roll.)
Similarly, going back to Allan’s comment, if you want to be well-supplied enough that you can go through three strikes… Well, then you should have bought three lots of bullets. If you didn’t bring three lots of bullets, then you WERE undersupplied.
A bit late to the party, but I do see a problem here. I buy two quivers of arrows, and I get two attack rolls. There is no way I’d run out that quickly.
A torch is even worse. I light it and get a 1 the very first roll.
I like the variance, and the reduced tracking.
So definitely need the stepping down option, so starting at a say a d12 for each quiver, I’d get at least seven rounds of attacking.
Another option, such as for a torch, is to have 2d4. If either or both roll a 1, then you go to one torch, and when that rolls a one, then the torch is finished. This gives a minimum of two turns.
So here is my suggested system.
Ammo and other large quantity items are bought in packs (ie quivers for arrows), and each pack is a d12. Rolling a 1 reduces -> d10 -> d8 -> d6 -> d4 -> d2 -> 0.
Things like wands with unknown charges, the GM picks a die size, and when it rolls a 1, that is it. Players don’t know the remaining charges so it could run out at any time, so players shouldn’t see any degradation, it should be a surprise when it is out.
Things like durations, have three dice of one size. Do not roll the first time segment (should be easy enough to not need tracking). Roll one die each time segment after, and when it rolls a 1, then you roll 2 dice each segment, if either rolls a 1, that time segment is the last. Roll at the beginning of a time segment, and on the final segment the character can often tell they are about out of time. The torch is almost to the hand and growing feeble, the mage armor is looking a little thin and less vibrant, etc.